Targeted metabolomics reveals multiple metabolite alterations in the urine of transition dairy cows preceding the incidence of lameness

Abstract Text:

Lameness (Lam) is a major issue of transition dairy cows consuming high grain diets affecting 25-35% of the herd. It is associated with decreased milk production, fertility problems, and high culling rates and treatment costs. Various hypotheses have been forwarded during the years with regards to the causes of lameness including rumen histamine, endotoxin, or biogenic amines. Although much is known about non-mechanical lameness, the precise pathobiology is not known. The objectives of this study were to evaluate weekly metabolite composition of urine in dairy cows starting from the beginning of dry off until 8 wks postpartum. Urine samples were collected from 100 cows at -8, -4, disease week, +4, and +8 wks around calving and stored at -80 C until analyzes. DI/LC-MS/MS analyzes were conducted on samples collected from 20 healthy control cows (CON) and 6 cows diagnosed only with lameness (no other periparturient diseases). A total of 154 metabolites including 41 carnitines, 9 lysophosphatidylcholines, 74 phosphatidylcholines, 15 sphingomyelines, 11 amino acids, 2 biogenic amines, hexose, and carnosine were identified and quantified. Data were processed statistically by MetaboAnalyst and univariate analyses. Results showed that 41, 29, 59, 26, and 40 metabolites were identified and measured to be different between the two groups on -8, -4, disease week, +4, and +8 wks around calving, respectively. The highest number of altered metabolites was identified during the week of diagnosis of Lam. Several metabolic pathways were found to be associated with the disease including amino acid metabolism, catecholamine biosynthesis, protein biosynthesis and urea cycle at -8 wks precalving; cysteine, glutamate, tyrosine, and glutathione metabolism at -4 wks precalving; and beta-alanine metabolism during the week of disease. ROC analyses, for determination of specificity and sensitivity of potential biomarkers, identified several metabolites with AUCs for ROC curves 0.96 (95% CI, 075-1.0) at -8 wks; 0.971 (95% CI, 0.905-1.0) at -4wks; 1.0 (95% CI, 1.0-1.0) at disease week; 1.0 (95% CI, 1.0-1.0) at +4wks; and 1.0 (95% CI, 1.0-1.0) at +8wks postpartum. PLS-DA analysis also showed clear separation between the two groups of cows with regards to altered metabolites. In conclusion, targeted metabolomics can be used to identify metabolic alterations in the urine of dairy cows before, during, and after diagnosis of Lam; it can also help to better understand the pathobiology of disease, identify cows more susceptible to Lam, and develop new preventive strategies.

Keywords: Dairy cows, DI-/LC-MS/MS, lameness, targeted metabolomics, urine